Veritula – Meta
Showing only #3476 and its comments.
See full discussionLog in or sign up to participate in this discussion.
With an account, you can revise, criticize, and comment on ideas.
Feature idea: pay people to criticize your idea.
You submit an idea with a ‘criticism bounty’ of ten bucks per pending criticism received, say.
The amount should be arbitrarily customizable (while covering transaction costs).
There could then be a page for bounties at /bounties. And a page listing a user’s bounties at /:username/bounties.
When starting a bounty, the user indicates terms such as what kinds of criticism they want. This way, they avoid having to pay people pointing out typos, say.
Anyone can start a bounty on any idea. There can only be one bounty per idea at a time.
How do you ensure the criticism is worthy of the bounty?
There could be a grace period. For example, 24 hours after the bounty ends, no new criticisms can be posted on the bountied idea. That way, the bounty initiator has time to review pending criticisms.
What happens if they neglect to review pending criticisms?
Then all pending criticisms automatically receive equal payouts.
I’m not sure yet, but I’m playing with the idea that the criticism can’t have any pending counter-criticisms by some deadline. Each counter-criticism could reset the deadline to give everyone ample time to respond.
The bounty initiator’s card will have to be authorized when starting the bounty. Card authorizations presumably have a deadline, so resetting the deadline won’t be an option.
Couldn’t I let the initial authorization expire and then re-authorize the card?
Maybe, but what if re-authorization fails? Then nobody gets paid.
What if Veritula charges the card immediately and holds the funds?
That seems like a tough sell. Users might not be willing to spend money without knowing whether anyone will submit any criticisms.
What if every criticism on the bountied idea creates a separate authorization? The bounty initiator would add their card on file and then every time someone submits a criticism, the card is authorized for the per-criticism amount.
While this idea sounded promising at first, I now realize it just moves the deadline problem one level underneath the bountied idea.
This idea introduces additional complexity and edge cases. For example, what happens if authorization fails? Need something simpler for an MVP version of this feature.
But then bad actors could always submit arbitrary counter-criticisms just before the deadline to avoid paying.
Idea: when you create a bounty, you set the amount you’re willing to pay per criticism and a ceiling for the total you’re willing to spend (no. of crits * amount per crit).
Your card is authorized for twice the ceiling. In addition, there’s a button to report abuse. If you’re a good citizen, you’ll be charged the ceiling, at most. But if you’re found to submit arbitrary criticisms to avoid paying, the bounty stops early and your card is charged the full authorization.
That doesn’t address the possibility of others submitting arbitrary criticisms just before the deadline.
What incentive would others have to submit arbitrary criticisms? They’re not the ones paying.
People who have submitted criticisms hope to get paid. They have an incentive to submit arbitrary counter-criticisms to others’ criticisms to increase their own share of the pie.
Although there’s a risk for abuse, that’s a feature: it will lead to lively discussions among critics.
There could still be a button to report abuse. People found to abuse deadlines could become ineligible for payouts and excluded from participating in future bounties.
The timeframe to address the criticism should start counting down from the moment the criticism is made, rather than the original post. So it would be a continuous thing rather than a single deadline for everyone.
The OP could end the bounty if there are no outstanding criticisms and he no longer seeks a solution.
I suppose that would make it a bit harder for bad actors because they’d need to monitor multiple deadlines, but they could still submit arbitrary counter-criticisms just in time to avoid paying. Or is there something I’m missing?
I would have it that each criticism and counter-criticism resets the countdown on the bounty deadline. This means everyone involved is given fair time to respond at each turn.
If you submit a criticism, you won’t want to wait indefinitely to get paid just because others are keeping the discussion going in a different branch.
Sorry but I don’t see how that solves the bad-actor problem. Bad actors would still be able to draw out the discussion to avoid paying, wouldn’t they?
Yes, but bad actors are a separate problem to solve (as you have alluded to in #2513, where you mention “good citizens”).
The problem we are addressing here is giving people a fair time window to respond to criticisms after they are published.
Edit: after reviewing the thread, I see that you were more focused on the bad actors problem while I was more focused on giving people fair time to respond. I believe what I am saying still stands, but maybe it belongs somewhere higher up the chain.
Card authorizations will necessarily have a deadline.
That is simply an extension of the fact that solutions to problems don’t come reliably.
Yes, that was what I was thinking. Presumably the OP could set their own deadline timeframe too.
Rather than set a fixed amount for each pending criticism (#3421), the ceiling could be divided among all pending criticisms equally.
But that would mean that the first criticism receives a payout at the same time the last criticism receives a payout. That creates an incentive to ignore new bounties in favor of older ones.
There is a counter-incentive to be the first to submit a criticism since subsequent criticisms run the risk of being duplicates, and being criticized as such.
Given the need for a deadline, all critics get paid at the same time anyway.
That could result in amounts too small to cover transaction costs.
Could pay out to only first x criticisms, where x is small enough the payout for each criticism is high enough to cover transaction costs (and then some).
Then it’s less clear to contributors how much money they can expect.
There could be a UI component showing estimated payout based on current number of criticisms, with a warning that actual payout could be less.
Let’s say somebody starts a bounty with permissive terms, asking for virtually any kind of criticism. They set a high ceiling, hoping for many submissions. $200, say.
If they only end up getting one or two small criticisms, for typos, say, they won’t like having to pay 100 bucks a pop.
In other words, the few criticisms you end up getting may not be worth the ceiling.
In a future iteration, the user could additionally set a per-criticism ceiling. Which the site would recommend setting when using permissive terms.
This way, the user could set a total budget of $200, say, while capping each criticism at $30, for example. The first 6 eligible criticisms would each get $30, and the next one would get $20. The remaining criticisms would get nothing.
This approach effectively merges #3474 and #3472, giving users maximum flexibility to choose the best outcome depending on what kinds of criticism they anticipate getting based on their terms.
Could this feature be unified with #2669 somehow?
The initiator of the bounty could choose a ceiling for the total they are willing to spend. They could additionally specify the amount per pending criticism.
For example, a user would indicate that they are willing to spend a total of $100 at $10 per criticism.
This approach is more complex for the bounty initiator than just indicating a total amount they are willing to spend (#3474). It’s best not to require users to do math.
But that means that additional criticisms don’t get any payout.
People can speculate, but if there’s a grace period, they won’t know how many pending criticisms there will be in the end. So they may still get a payout.
That could be a good thing in that people won’t completely overwhelm OP with criticisms.
Unlike #3424, however, having a set amount per pending criticism means there’s zero incentive for anyone to submit more criticisms, whereas divvying up the amount among pending criticisms means the incentive is reduced only gradually, and it’s up to people to decide for themselves whether contributions are still worth making.